Faculteit der Economische Wetenschappen en Bedrijfskunde
SERIE RESEARCH MEMORANDA
Coping with Uncertainty
An Expedition in the Field of New Transport Technology
Marina van CeenhuizenPeter Nijkamp
Research Memorandum 2001-9
March 2001
vrije Universiteit amsterdam
COPING WITH UNCERTAINTY
An Expedition in the Field of New Transport Technology
Marina van Geenhuizen *
Peter Nijkamp * *
Abstract
Many decisions of mankind are rational only to a limited extent. This holds for individualtravel behaviour, but also for long-range strategic decisions on transport systems or transporttechnology. In any decision problem coping with uncertainty is the most critical element.The introduction of new transport technology is surrounded by many types of uncertainty. Forexample, there is uncertainty about the pace and extent of adoption of new technology andthere is uncertainty about the impact of new technology in terms of an increased sustainabilityor increased efficiency. This article attempts to map uncertainty surrounding new transporttechnology and to identify ways to deal with uncertainty in policy making. The findings willbe illustrated with electric vehicles, particularly with two specific strategies to deal withuncertainty, i.e. interactive technology watching and experimentation in a market niche. Thearticle concludes with a discussion of success factors that influence the outcomes of suchstrategies.
*Faculty of Technology, Policy and Management, Delft University of Technology, Delft**Department of Spatial Economics, Free University, Amsterdam
Pn893mvg
1
t
1. Homo Mobilis and Homo Economicus
Conventional wisdom in transport science suggests that transport movements serve to
overcome physical distance barriers. Such movements are - apart from a few rare cases like
leisure trips without a destination - necessary because of the geographical separation of
various activities (residential, employment, social etc.). Since transport is regarded as a
burden, the rational decision-making paradigm then suggests that it is wise to minimise the
transport effort. In particular, welfare theory - one of the comer stones of transport economics
- argues that it is in the interest of the ‘homo economicus’ to minimise travel costs
(pecuniary, time-wise or psychological).
This assumption lies at the basis of many utility-based demand studies (e.g., discrete choice
models) but has also created the foundation of many aggregate transport studies (such as
quadratic assignment models, linear programming models for modal split or route choice and
the like). In fact, the overwhelming majority of transport studies looks into physical
movement as a sacrifice. If this basic assumption were valid, then a series of intricate and
intriguing questions emerges. How come that the action radius of modem men tends to
increase structurally (see, e.g., van Doren, 1991)? How is it possible that the ‘law of
conservation of travel time’ is rather robust (at least in metropolitan areas), but shows an
increasing trend for longer trips? Is the modem ‘homo mobilis’ essentially a ‘homo
economicus’ or is he driven by other - also rational - instincts?
In a recent study on ‘slow motion’ (see Nijkamp and Baaijens, 1999) a review of motives in
favour of other - less fast and (perhaps) more environmentally benign - modes of transport
was given, based on the recently developed concept of a ‘time pioneer’. The argument is that
it may - from both an individual and a collective welfare perspective - be rational to reduce
travel speed. Empirical research however, demonstrated that the willingness among travellers
to choose ‘slow motion’ options is disappointingly low. On the contrary, the trend is towards
more speed and a larger action radius. Does this mean that the ‘homo mobilis’ is not a rational
decision-maker? It should be noted that the basic hypothesis in welfare theory is that the.individual seeks to maximise his (her) own welfare, based on the assumption of perfect
information and on the absence of externalities, bandwagon and image effects, and the like.
There is quite some literature, in particular in the 1960s geography, which convincingly
demonstrates - both conceptually and empirically - that the choice of the actual location of
2
firms can differ significantly from the theoretical optimal location (see e.g., Wolpert, 1964).
This is then explained by the prohibitive role of high search costs to find the optimum
optimorum. Choice makers tend to be often satisfied with a second-best - or even third-best -
location, as they are not driven by the exclusive goals of profit seeking behaviour. The
theoretical basis of this ‘satisficing’ behaviour has been given by Herbert Simon in this theory
on ‘bounded rationality’ (see Simon 1952).
Apart from the costs involved in gathering perfect information, there is also a social driving
force. Location and travel decisions are not taken in a Robinson Crusoe economy on an
isolated island, but in social interaction with others (the notion of the ‘homo socialis’). For
example, the need to live near one’s relatives may lead to a residential location decision that is
not optimal from a cost-minimising viewpoint. Consequently, there is quite some evidence
that travel decisions are not exclusively governed by rational - cost-minimising, utility-
maximising or profit-maximising - behaviour. Theoretically speaking, this would imply a
specification bias in many of our behaviour transport models. In any case, the ‘homo mobilis’
is faced with many uncertainties in his (her) travel choice; perfect information is an illusion.
This recognition may have serious implications for the precise value of transport models and
limits seriously the scope of cybernetics oriented engineering approaches in transport science.
There is still another issue. To assemble proper information may be too costly, but the human
ability to digest all relevant information is also limited. Consequently, the available or
existing information may also be selective and has to be interpreted from the subjective
perspective of the user or decision-maker. Furthermore, if travel information has to be
transmitted to road users, the question is whether a particular road user needs exactly the
given information. In many cases, he may be willing to receive additional or other
information. This is clearly reflected in an early statement of March and Simon: “...the vast
bulk of our knowledge offact is not gained through direct perception but through the second-
hand, third-hand, and nth-hand reports of the perceptions of others, transmitted through the
channels of social communication. Since these,perceptions have already been filtered by one
or more communicators, most of whom have frames of reference similar to our own, theareports are generally consonant without the filtered reports of our own perceptions, and serve
to reinforce the latter” (March and Simon, 1958, pp. 153).
3
The previous remarks do not only apply to daily or routine decisions. They are equally
relevant for strategic decisions of a long-range nature on transport systems or transport
technology. Coping with uncertainty is essentially the most critical element in any decision
problem. This question extends far beyond stochastic or probabilistic approaches to
uncertainty management. In the sequel of this article we will address in particular uncertainty
issues in the field of transport systems technology, but most remarks apply also to uncertainty
problems in the broader area of transport management and policy.
2. Exploring the Future
In the past decades we have observed a great variety of attempts to come to grips with an
uncertain future. This is clearly exemplified by the path breaking work of Kahn and Wiener
on “The Year 2000”, published in 1967. The question is whether much progress has been
made in the past decades. This provokes in particular a complicated methodological issue, as
it asks whether better understanding does also lead to better forecasting. Some remarks on this
issue are in order now. Explanation is a process of logical deduction and empirical validation,
in particular in the context of repetitive events. In a strict sense it would be impossible to
subject “unique” events to a scientific explanation. In any case, an explanation in a
probabilistic statement of an “if-then” nature, which - given a predetermined domain - is able
to make a scientifically founded statement on the possible or probabilistic occurrence of
events as part or as a result of events which fall outside the a prior, defined domain. Thus, an
explanation has quite some restrictions. Its validity is limited by prior hypotheses and by
methodological constraints.
From a methodological perspective there is not so much distinction between explanation and
forecasting, apart from an obvious difference between prospective and retrospective thinking.
But also forecasting is a process of logical analysis on the basis of empirically verifiable
relationships, with the only exception that exogenous variables - falling outside of the
forecasting domain - cannot be empirically observed, but can at least be hypothesised (e.g. on.the basis of plausible reasoning). But in principle one might argue that forecasting =
explanation.
In the history of future studies we may distinguish various approaches which aim to offer a
scientific underpinning for statements on future events. Examples are the following:
Blueprint thinking. In this mode of thinking the basic assumption is that it is possible to
define a feasible amount of targets in the future, which can be realised through the right
use of instruments. The idea of a “makeable” society - very popular in the seventies and
eighties- is illustrative for this approach.
Normative thinking. Here the idea is that a society may evolve according to a priori fixed,
normative -sometimes ethical- criteria.
Nested thinking. This is based on the conviction that the future as a whole is difficult to
forecast, but that certain parts or aspects (e.g., demography, technology) may be mapped
out with a sufficient accuracy.
Fiction thinking. This type of future thinking is based on a non-historical approach and
questions the limits posed by past constraints.
Scenario thinking. This has become a popular way of analysing the future. It presupposes
the design or existence of meaningful - not necessarily realistic - future images, but it
demands that such images meet strict methodological requirements such as internal
consistency and a bridging between the present and the future. Such scenarios are
politically not committing, but offer a spectrum of future developments (e.g. Nijkamp and
van Geenhuizen, 1997; Nijkamp et al., 1998).
Evolutionary thinking. This approach is not based on normative or policy views, but
assumes some sort of adjustment mechanism in human behaviour (e.g., resilience), which
drives living organisms or systems towards “continuity in change”. It also draws attention
to a phenomenon named path dependency, meaning that future moves in policy making
are constrained by decisions made in the past.
Learning thinking. This mode of thinking has already a long history (see, e.g. Harvey,
1967) but has recently become fashionable in connection with evolutionary thinking.
Popular concepts in this framework are the learning decision-maker, the learning
organisation, the learning city, etc. This principle forms a contrast with blueprint and
normative thinking and avoids “linear” thinking. It is based on a blend of positive and
negative feedback responses in a dynamic choice environment (e.g. Stacey, 1992).
5
The previous observations hold also for transport technology. The high risks of new
technologies may hamper a rapid introduction. But also the acceptance attitude of the user
may cause a stumbling block. These issues will be further discussed in the next sections.
3. The Uncertain Roads of Transport Systems
Transport is the spatial projection of societal dynamics. The 21” century will see a variety of
new social and technological trends which will influence the way in which transport is
supplied and utilised (cf. Becker, 1997). At present a wide range of social phenomena,
including rising incomes, increased leisure time, new communication technologies, an ageing
population and a declining role for the traditional family, are changing the nature of the
demands we place on transport. In response to new techniques of production, shipping and the
growth of markets, economic activities are also changing. Institutional reforms such as
privatisation and deregulation have changed transport in ways that are not yet well
understood. At the same time increasing use of petroleum resources for travel and transport
has raised concerns about the eventual depletion of fossil fuels as well as its contribution to
global warming and decreases in urban air quality. But from a methodological perspective we
have to ask ourselves: do we have the scientific apparatus to forecast uncertain future
developments, trends or events, particularly those that are crucial in the adoption of new
transport technology?
New technology may affect the transport system in two different ways. First, there is
optimisation of the current transport system by using existing potentials, and secondly, there
is the much more comprehensive structural change of the transport system. A good example
of the first is the electric vehicle because the concept of individual cars remains unaffected in
merely challenging the type of traction system. However, if electric vehicles are used in new
concepts of carpooling and mixed forms of private and public transport (using busses)
attention moves to structural changes in the transport system (Rutten et al., 1998).
Accordingly, the latter case involves much more uncertainty.
.
Uncertainty can be tackled from various angles. For example, Rowe (1994) in elaborating
uncertainty as the “absence of information”, makes a division into various dimensions (van
Geenhuizen et al., 1998):
6
l Temporal uncertainty. This refers to the future state of the transport system (prediction
uncertainty) and related systems, but also to the past states of these systems.
l Structural uncertainty. This refers to the complexity of the transport system in modelling.
Structural uncertainty depends on the number of parameters and interaction used in
models to describe a situation or development. The key question is how well the models
represent reality.
l Metrical uncertainty. This is concerned with difficulties in measurement of the system.
Central issues here are precision and accuracy (validity). Such uncertainty plays a role for
example, in measuring preferences of travellers in public transport.
l Translational uncertainty. This refers to the communication of the results of analysis and
modelling in a policy context. Goals and values of the stakeholders are the main issue
because these affect the interpretation of the results and the related effort in consensus
building.
All four dimensions of uncertainty occur in any situation and reinforce each other. The last
dimension - translational uncertainty - indicates the importance of stakeholders in policy
making. Their interests may be different, dependent upon their goals and means, problem
perception and interpretations. An interesting example in this framework is the estimation of
future traffic flow and the estimation of costs of large transport projects (Skamris and
Flyvbjerg, 1997). A compilation of before-and-after-studies of various large projects shows
considerable cost overruns (of 50-100%) and traffic forecasts that are remarkably optimistic
(by 20 to 60%). The differences between forecast and actual costs and traffic cannot be
explained primarily by structural uncertainty. By being strongly consistent and one-sided the
uncertainty that has entered is much more of the translational type and needs to be seen - as
Wynne (1992) puts it - as a “social construct”.
By adopting the perspective of policy making and by conceptualising the relevant part of
reality as a system, four classes of uncertainty can be distinguished (see the introduction of
this special issue):
a
l Uncertainty about the future outcomes of the system
l Uncertainty about the future behaviour of external influences
7
l Uncertainty about system behaviour, connected with the above mentioned structural and
metrical uncertainty
l Uncertainty in the selection ofpolicy measures, including guiding values.
This classification serves as a framework in our identification of uncertainty in policy making
on the adoption of electric vehicles in the next section.
4. Policy Making concerning Electric Vehicles
The development of electric vehicle technology during the end of the lgti and beginning of
the 20” century exemplifies highly dynamic processes with unexpected outcomes (Isoard,
1996; Cowan and Hulten, 1996). At the start of the market introduction at that time electric
cars were far ahead compared with gasoline cars. The gasoline engine was considered to be
inferior because of the noise it produced, and its unsafe and dirty character. There was also a
need for some additional but complicated technical innovations. However, the strategic
behaviour of a few gasoline car producers - especially their move to a mass market and use of
specific marketing tools - has given the gasoline car a position ahead. Due to subsequent
advantages from increasing returns to scale (and lock-in), the gasoline engine could maintain
this position whereas competing technologies faced progressively greater difficulty in
entering or surviving in the market. This example shows the ill-predictability of the winning
technology - connected with various small events (or actions) - and the non-superiority of the
selected path.
A detailed analysis of the adoption of electric vehicles as it is to-day and of concomitant
policy making brings a differentiated “landscape of uncertainty” to light (van Geenhuizen and
Schoonman, 1999; Mulder et al. 1996; Rutten et al., 1998). The key uncertainty belongs to the
behaviour of the system, i.e. user acceptance, and to the future outcomes of the system (Table
1). Accordingly, it is difficult to select the appropriate policy measures. In fact, we are not
able-to model user acceptance behaviour and policy impacts in such a way that it helps to
predict the adoption of electric vehicles and to get insight into underlying factors. Metrical
uncertainty, of course, contributes to the lack of adequate prediction results. For example,
despite progress there are still measurement problems to solve in studies focusing on stated
preference (of users) (e.g. Golob et al., 1996) and on opinions (of experts) (e.g. Mulder et al.,
8
.‘.
1996). The uncertainty that stems from the critical user acceptance seems connected with
performance and price aspects of electric vehicles, i.e. a relatively small range, danger of dead
battery and relatively expensive to buy, but there is more. Over the past decades cars have
increased freedom, mobility, prestige, and a whole lot more in the life of individuals (e.g.,
Marsh and Collet, 1986). It may take therefore, a number of years to add the totally different
attribute of environmental ideals to this range. Moreover, electric cars may never attract
drivers that are moved by car appeal going far beyond merely functional transport.
Knowledge of such appeals and their influence is too small to date in order to contribute to
sound predictions. The same holds for uncertainty in the case of mixed private-public
transport using electric busses. Here, we may expect user resistance based on a general
aversion against all modes that are different from the individually used car. On the other hand,
environmental motives may become strong in the segment of second cars in households and
in the segment of business cars. Both are growth markets, and we may be able to predict how
fast these segments will grow in the next coming years.
Table 1 Uncertainty in policy making for electric vehicles
Element Details
Future outcomes
External influences
System behaviour
Selection of policymeasures
*
- Achieving aimed results- Causing negative side-effects- Behaviour of actors in dominant technologies
(resistance from lock-in), such as conventionalcar manufacturers and fuelling industries
- Growth of customer segments with largeopportunities (second cars, business cars)
- The time that high performance batteries andfuel cells are available at a competitive price
- Divides within the technology, such as low andhigh temperature battery systems, leading topotential incompatibilities
- Critical user acceptance: Willingness to payextra and bear inconvenience for environmentalgains; willingness to “lose freedom orindependence” in a shift to non-private transport
- Effectiveness of measures, e.g. of incentives,subsidies
- Missing guiding values
Strength ofuncertaintyStrongStrong
Weak
Weak
W e a k
W e a k
Strong
StrongWeak/strong
The uncertainty in future outcomes of the system can be summarised as follows. Due to
uncertain system behaviour and uncertain inputs, the latter mainly from external factors, we
9
are not able to predict the outcomes of the system in terms of a move towards reaching policy
aims, such as sustainable energy aims and prevention of externalities (noise, emission). The
same holds for adverse effect, like an increase in overall car mobility, due to the use of
electric cars as second (third) cars in households. Buying electric cars may serve as a kind of
justification, whereas without a pushing policy of electric cars no second or third cars would
have been bought.
As far as technology is concerned, uncertainty seems far less strong. Battery and fuel cell
technology will need some more 5 years before these can be used in high performance and
cost competitive electric vehicles. Research on improved performance and reduction of
production costs of light weight rechargeable solid-state lithium-ion batteries is of particular
interest. Research focuses on the stability, compatibility, and electrical properties of the
battery components. World-wide research also focuses on cost reduction of polymer fuel
cells, among others by reducing the amount of platinum catalyst. The uncertainty that is
involved seems to be “manageable” in that it can be reduced by increased research and
development efforts (van Geenhuizen and Schoonman, 1999).
In terms of guiding values, an important missing link in policy making to date is a clear vision
on the functions of transport, other than the straightforward bridging of distance. These other
functions may be strongly (social) psychological, like a gain of individual freedom, increase
of power, and gain of identity. In a situation in which it is generally accepted that car driving
only serves the functional bridging of distance, designing a policy to enhance the adoption of
electric vehicles is less difficult compared with a situation without such a guiding value.
The above circumstances indicate a large uncertainty in the selection of policy measures to
enhance the adoption of electric vehicles. The way how policy makers may deal with
uncertainty will be discussed in the next section.
5. Strategies to Cope with Uncertainty.
There are many strategies for coping with uncertainties, each using different methods. The
following strategies can be distinguished (e.g., Friend and Hickling, 1997; van Geenhuizen et
al., 1998; van Geenhuizen and Thissen, 2001):
1 0
t
l To ignore uncertainty, take policy measures, and see what will happen; while this may be
the easiest option for the short term, it means in fact accepting large uncertainty with
respect to the policy outcomes.
l To identify and, if possible, specify uncertainty. This enables the policy maker to act
consciously in the presence of uncertainty. The classifications above may be helpful in
this respect. Methods to identify and specify uncertainties include scenario analysis (e.g.
for uncertainty in the system’s surroundings) (Nijkamp et al., 1998; von Reibnitz, 1998;
Schwartz, 1991), development of alternative system models to account for alternative
structures, and determination of confidence bounds for data and model parameters.
l To do nothing and let uncertainty be reduced by time; this approach means delay, and
involves a good chance that, while some uncertainties will disappear, new ones will
emerge. Doing nothing may be a positive choice, based on a vision in which the self-
organising capacity of society and particular interest groups is given priority and is trusted
to achieve satisfactory outcomes.
l To reduce uncertainty. This can be done in different ways. First, by additional research or
better integration of existing knowledge; this might cover the range from the collection of
new data to the application of advanced methods of integrated modelling and explorative
modelling to more clearly distinguish between possible and impossible developments, and
to identify critical events and bottlenecks. Second, by pushing the uncertainty onto
someone else, which generally will involve costs (e.g., insurance premiums,
compensation). Third, by negotiating with others whose behaviour is uncertain but has a
significant impact on the desired policy outcomes. This may involve processes of
participatory policy development, and/or agreements with policy makers in adjacent
fields.
l To accept uncertainty, and act consciously in its presence. Here, too, different strategies
are possible. A robust policy may be selected, i.e., a policy that will do well in most
possible future circumstances. Or a flexible policy is designed, i.e., a policy that is
adaptable to the future course of events (e.g. Walker et al., 2001).
l To see uncertainty not as a threat, but rather as an opportunity to creatively shape the
future. Rather than emphasising a choice for a presently available policy option, this
approach calls for development of a vision that provides the guiding principles for present
and future action, such as experiments and other learning exercises that may underpin
policy measures (Stacey, 1992).
1 1
:‘.
‘I
I’. _ _-
It seems self-understanding that the above strategies do not exclude each other but often
complement each other. We now turn to two complementary roles adopted by the Dutch
government in an attempt to identify and specify uncertainty, and learn about the future by
experimentation, i.e. the role of technology watcher and the role of innovation agent (Grin et
al., 1997; Rutten et al., 1998).
The role of technology watcher includes a systematic monitoring of technological
developments and understanding of these developments in terms of repercussions for policy.
The developments are positioned in a frame of five layers which provides the appropriate
background to assess the relevance of the new technology in transport. These layers include
the one of components (such as batteries and fuel cells in the case of electric vehicles), the
application (e.g. concepts of propulsion), the type of vehicle (cars, busses), the transport
concept (short or long distance transport), and the transport system (such as multi-modal
commuter transport systems). The main activities of technology watching are:
(1) to collect useful information on a particular technology and its potential markets
(2) to arrive at a broad interpretation of this information leading to score matrices for the
technology on specific criteria
(3) to establish various rounds of interaction with policy makers, policy analysts and
researchers in order to focus on policy relevance
(4) to establish a synthesis and preliminary conclusions
(5) to distribute the results to policy makers and other stakeholders.
What is new in this approach is the acknowledgement of the fact that each of the above
activities includes a number of choices, for which a solid and transparent argumentation needs
to be provided preferably in interaction with important stakeholders. Apart from the selection
of the technology itself, there is for example the selection of the type of information to be
collected (activity 1) and the criteria to be used in a first interpretation (activity 2). The key
activity of course, is to scan the technology for policy relevunce (activity 3). This includes
both-an assessment of direct impacts and an exploration of (potential) use of the technology
according to various criteria. The latter activities may include various rounds of scenario
analysis in order to learn about the technology. In terms of results, the methodology may not
only clarify the potential role of the technology but also the role of factors that advance or
prohibit this role.
1 2
It is increasingly acknowledged that a participatory (or interactive) structure is a key
prerequisite for success of the methodology. This means a continuous input from the side of
independent experts and users of the technology at hand. There is a trend to consider a broad
group of stakeholders as relevant, i.e. also including actors on the supply side of the new
technology, like manufacturers and sponsors, and organisations that contribute to the
embedding of the technology (Grin et al., 1997). This broad input serves to cover the
perspectives of all influential players at hand and, accordingly, to increase support of them.
But there is also a danger here. It may happen that results are achieved without sufficient
reflection of real-life power structures, thus being unrealistic in a sense. This calls for a
delicate design of participation to arrive at strong outcomes that have sufficient “authority”.
As previously indicated, in its role of innovation agent the government is actively involved in
a successful implementation of the new technology. One way of doing this is to experiment
with a new technology in a relatively small and protected niche of the market, and this is
exactly the policy for electric vehicles in the Netherlands. The major experiment deals with
daily commuting between the cities of Rotterdam and Zoetermeer, in which electric busses
are used to bring commuters to their work at the location of the Ministry of Transport and
Waterworks in Rotterdam. Such an experiment serves various important goals, like to
demonstrate the use of the technology and to test the technology on user value and
acceptance. Technology experiments in niches like this one, require the identification of the
right market niche, a smart combination of resources and the involvement and interaction of
the right stakeholders. More importantly, network management is involved in terms of
shaping and reshaping the relationships between stakeholders (de Bruijn and ten Heuvelhof,
2000). The ultimate goal of such experimental small scale implementation is to proceed in the
adoption itself, while learning about advancing and hampering factors. The outcomes then
serve as input for a further development and testing. Learning experiments like these match
with a planning culture that is relatively new, i.e. vision-based planning (Figure 1).
An essential difference with conventional ,planning approaches is that there are no
predetermined goals to be achieved by using new technology and no predetermined policy
measures that serve the realisation of these goals. Rather, there is a broad vision on new
technology and transport, and there is a continuous interaction (communication) with
stakeholders. Information from these two sources provides the basis for conducting various
1 3
experiments. Further, the learning derived from a series of experiments (trial and error) may
arrive at a stage in which it enables to underpin decision making.
Figure 1 Vision-based Planning
Events in the transport system
Ilisten tostakeholders outcome
trial and error
Source: van Zuylen, in van Geenhuizen et al. 1998.
Similar to the previous discussed role of technology watching, the experiments need to be
selected, designed and carried out in such a way that there is a broad coverage of causal
factors as well as a broad support from stakeholders for the results, without losing sense for
real-life power relations. Note that vision-based planning is an interesting alternative to
conventional practices only in specific cases. It is not regarded here as a general solution to
problems of uncertainty in transport policy making, for the simple reason that particular
policy questions are concerned with large scale infrastructure projects like high-speed rail
connections and Maglev systems. In these cases, experimentation in a niche and learning
based on trial and error are not useful.
6. Variation in Europe
There are different national planning methodologies in Europe. For example, in some
countries there is an emphasis on the input of users in technology assessment, whereas in
other countries experts have a prominent role. The importance attached to reduction of
uncertainty in policy making also varies from country to country, based on differences
1 4
between national cultures (de Jong, 1998; Stough and Rietveld, 1997; van Zuylen et al.,
1998b).
What roles governments actually play with regard to promising transport technology is very
much dependent upon a blend of factors. In policies for introduction of electric vehicles in
various European countries in the mid 199Os, we observe large differences that can hardly be
ascribed to differences in planning culture only (IAEA, 1995; Weber and Hoogma, 1998)
(Table 2). Rather, various meso- and macro economic factors seem to count. For example, the
Netherlands compares with Germany in the dominant type of electricity production. In both
countries, electric vehicles using batteries would be powered with electricity from non-
sustainable sources (coal and gas), a situation in which the environmental gains of the
technology can be questioned and the public image of electric vehicles seems relatively weak.
Note that this might change if hydrogen powered fuel cells are used.
Other important factors that differentiate between countries include price and availability of
electricity, and the match of electric vehicles with accepted problem situations such as the
recognition of air pollution in cities as a severe problem and general environmental problems.
Furthermore, the industry structure seems important with monopoly and oligopoly enabling a
smooth involvement of national governments. There is one factor of which the influence is
not quite clear, namely the structure of the government system. In Germany, a decentralised
system would hamper consensus building on the need for adoption of the new technology, due
to the involvement of different governments on a lower level. However, in the case of
Switzerland a similar government structure would not have a hampering influence, most
probably because the new technology is not questioned due to its zero emission character. The
Swiss government also acts as a conditioning agent in the establishment of a kind of “clean air
act”. According to this policy, a total of 200,000 electric vehicles are planned on the road in
2010, corresponding to about 8% of the number of cars in the country today (AssoVEL,
1998).
.
1 5
Table 2 Factors in macro (meso) systems and government roles in EV adoption
Country Factors Role of GovernmentFrance Advancing factors
Need for outlet for nuclear Subsidy for EV purchasepower Co-ordination of initiativesRecognition of air pollution in Funding of R&Dcities as a severe problemCentralised governmentIndustry monopoly andoligopoly
Sweden Advancing FactorsEV matches environmental Support of R&Dpolicy Support of demonstration andEV is nearly zero emission procurement programsLow price of electricity
Germany Prohibiting FactorsNon-renewable energy as basis Modest role (recentlyand full emission (coal firing) increasing)Decentralised governmentFragmented electricity industry
The Netherlands Prohibiting FactorsNon-renewable energy as basis Modest role:and full emission (gas firing) Support of R&DFragmented electricity industry Support of experiments in nichebut high co-operation (collective transport)
Switzerland AdvancingEV matches environmental Clean Air Actpolicy Co-ordination and subsidisationEV is zero emission of pilot programs
Subsidy for EV purchase
7. Concluding Remarks
Transport is a complex activity that touches upon many different human activities. As a result,
policy making is subject to uncertainty, a situation that calls for a strong awareness in the
design of policy measures. This article has mapped out the landscape of uncertainty in the
adoption of new technology in transport. In addition, it has considered various ways to deal
with uncertainty, in particular two new roles of the government in the Netherlands, i.e.
technology watching and experimentation in niche markets to advance adoption. What seems
1 6
, --.-
to be crucial for success of both strategies is the ability to create a sense of urgency among the
relevant stakeholders in involving them in the interaction at hand. At the same time, the
results need to mirror real-life power relationships between stakeholders. This serves not only
a large coverage in terms of influences, but also a sufficient support for and “authority”
assigned to the results. The latter is clearly necessary because the results of this type of new
strategies are often under-used by decision makers.
Moreover, one needs to be aware of the potential dynamics in external factors, meaning shifts
over time in stakeholders and upcoming new technologies. As we have seen in the early
history of electric vehicles, small events may cause a leading position of inferior technology
and a reinforcing of this position over time. The latter phenomenon would have two
implications for technology watching, i.e. to cover a relatively broad spectrum of technologies
and to identify the above indicated small but potentially very important events.
1 7
REFERENCES
AssoVEL Mendrisio, Large Scale jleet test with lightweight electric vehicle (LEVs) inMendrisio (2nd interim report). Mendrisio (Switzerland), 1998.
Becker, H., Social Impact Assessment, UCL Press, London, 1997.
Bruijn, J.A. de, and E. ten Heuvelhof, Networks and Decision Making. Utrecht, Lemma, 2000.
Cowan, R. and Hulten, S., Escaping Lock-In: The Case of the Electric Vehicle. TechnologicalForecasting and Social Change, 1996,56, pp. 61-79.
Doren, Ch. van, A History of Knowledge, Ballantine Books, New York, 1991.
Dror, Y., Uncertainty. Coping with it and Political Feasibility. In: H.J. Miser and E.S. Quade(eds) Handbook of Systems Analysis: Craft, Issues and Procedural Choices. 1988, pp. 247-281.
Friend, J. and A. Hickling, Planning under Pressure. The Strategic Choice Approach,Butterworth-Heinemann, Oxford, 1997.
Geenhuizen, M. van, Nijkamp, and H. van Zuylen (eds), Limits to Predictability in Trafficand Transport. AVV, Rotterdam, 1998.
Geenhuizen, M. van, and J. Schoonman, Sustainable Transport Technology: A Policy Viewon the Adoption of Electric Vehicles. Proceedings of CONAT ‘99, The ZXth ZnternationalConference. Transilvania University Press, Brasov, 1999.
Geenhuizen, M. van, W. Thissen, Uncertainty and the Adoption of Intelligent TransportSystems: A Policy View, Journal of Technology Policy and Management, 2001(forthcoming).
Golob, T. et al., Forecasting Electric Vehicle Ownership and Use in the California SouthCoast Air Basin. Institute of Transportation Studies, Irvine (CA), 1996.
Grin, J., van de Graaf. H. and R. Hoppe, Interactive Technology Assessment (in Dutch).Rathenau Instituut, The Hague, 1997.
IAEA (The International Energy Agency), Implementing Agreement for Electric VehicleTechnologies and Programs. IAEA, Paris, 1996.
Harvey, D., Models of the Evolution of Spatial Patterns in Human Geography, Models inGeography, R.J. Chorley and P. Haggett, eds., Methuen, London, 1967, pp. 549-608.
Isoarh, S., The Diffusion of Energy Technologies: The Influence of Increasing Returns. TheZPTS Report 1996 (10): 10-15.
Jong, W.M. de, Institutional Transplantation. Eburon Publishers, Delft (PhD thesis), 1998.
Kahn, H. and A.J. Wiener, The Year 2000. MacMillan, New York, 1967.
1 8
,.. ._ --
March, J.G. and H.A. Simon, Organizations. Wiley, New York, 1958.
Marsh, P. and Collet, P., Driving Passion. The Psychology of the Car. Jonathan Cape,London, 1986.
Mulder, K., Weijer, C. van de, and Marchau, V., Prospects for external sources of vehiclepropulsion. Results of a Delphi study, Futures, 1996, Vol. 28 (lo), pp. 919-945.
Nijkamp, P., and S. Baaijens, Time Pioneers and Travel Behaviour, Growth and Change, vol.30, no. 2, 1999, pp. 237-263.
Nijkamp, P., Ouwersloot, H. and S. Rienstra, Sustainable Urban Transport Systems: AnExpert-based Strategic Scenario Approach, Urban Studies, 1997, Vol. 34 (4), pp. 693-712.
Nijkamp, P. and M. van Geenhuizen, European transport: challenges and opportunities forfuture research and policies, Journal of Transport Geography, 1997,5, pp. 4-l 1.
Nijkamp, P., S. Rienstra and J. Vleugel, Transportation Planning and The Future, JohnWiley, Chichester/New York, 1998.
Reibnitz, U. von, Scenario Techniques. McGraw Hill, New York, 1998.
Rowe, W.D., Understanding Uncertainty, Risk Analysis, 1994, Vol. 14 (5), pp. 743-750.
Rutten, B.J.C.M., Korver, W. and Elzen, B., The importance of monitoring technology andexperimentation (in Dutch) Tijdschrijl Vervoerswetenschup, 1998 (3): 191-211.
Schwartz, P. The Art of the Long View: Planning for the Future in an Uncertain World.Doubleday, New York, 1991.
Simon, H., A Behavioural Model of Rational Choice, Quarterly Journal of Economics, vol.69, 1952, pp. 9-l 18.
Skamris, M.K. and Flyvbjerg, B., Inaccuracy of traffic forecasts and cost estimates on largetransport projects. Transport Policy, 1997,4 (3): 141-146.
Stacey, R.D., Managing the Unknowable, Jossey Bass, San Francisco, 1992.
Stough. R. and P. Rietveld, Institutional Issues in Transport Systems. Transport Geography,1997, 5(30): pp. 207-214.
Walker, W.A., S. Adnan Rahman, and J. Cave, Adaptive Policies, policy analysis, andpolicy-making, European Journal of Operational Research, 128 (2001): pp. 282-289.
.Weber, M. and R. Hoogma, Beyond National and Technological Styles of InnovationDiffusion: A Dynamic Perspective on Cases from the Energy and Transport Sectors,Technology Analysis & Strategic Management, 1998, Vol. 10 (4) pp.545-566.
1 9
Wolpert, J., The Decision Process in a Spatial Context, Annals of the Association of AmericanGeographers, vol. 34, 1964, pp. 537-558.
Wynne, B. Uncertainty and Environmental Learning, Global Environmental Change, 2, 1992,pp.1 11-127.
Zuylen, H.J. and van Schaick, G.H.J., Towards a technology policy in traffic and transport (inDutch), Tijdschrif Vervoerswetenschap 1998a, (3): 247-265.
Zuylen, H.J., Weber, M., Moon, D.P. and Davison, P., A new European approach totechnology policy in traffic and transport (in Dutch). Tzjdschrif Vervoerswetenschap, 1998b:267-287.
.
2 0
